-
Notifications
You must be signed in to change notification settings - Fork 7
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Timeline of loadgen stats #38
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM! I have just a few suggestions, but you're welcome to rebut them if you don't find them useful.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for catching those. I agree, splitting long lines is more readable. I need a prettier for bash, I've become to lazy to format my lines manually.
194b9cd
to
a2e2263
Compare
b2b3005
to
dbf3f59
Compare
a2e2263
to
9c0e3bd
Compare
4961771
to
afdc022
Compare
9c0e3bd
to
2541e2f
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nice
58cd8be
to
ceb913f
Compare
2541e2f
to
8717013
Compare
@michaelfig since your initial review I addressed your feedback and added d30dea3 and 8717013. Since those are minor changes, and @warner approved the PR following them, I will squash the feedback commits, and merge. |
8717013
to
bfe276b
Compare
Reorder CSV stats
This PR provides the plumbing to generate a CSV file containing a small subset of the loadgen stats across multiple revisions.
The file is currently generated locally and manually imported into a google spreadsheet. I still want to figure out a better way, by uploading the results automatically somewhere, and make them available to a hosted dashboard, possibly through a "revision completion" hook for the background script.
I generated a preliminary corpus of data on my own machine, with the results available here, using the following config:
--no-stage.save-storage --monitor-interval=1 --stage.duration=10 --stage.loadgen.vault.interval=30 --stage.loadgen.amm.interval=30 --stage.loadgen.amm.wait=15 --stage.loadgen.vault.limit=2 --stage.loadgen.amm.limit=2
A longer version is currently being generated on the benchmark machine for the same revisions with the following config:
--no-stage.save-storage --monitor-interval=1 --stage.duration=30 --stage.loadgen.vault.interval=60 --stage.loadgen.amm.interval=60 --stage.loadgen.amm.wait=30
I'd ultimately like to run the default 24h test in the background on a subset of past revisions to backfill the data, but I'm not sure which revisions would be the best candidates yet.
Stack on top of compatibility fixes (#37 / #44) and stats work (#36).
As always, best reviewed commit-by-commit.